After talking with the Exoslide guy and changing my hot end mount to give me space for a fourth slider on the hot end mount. I think that I have solved the problem.
My issue always occurred during a high speed transition in the X axis, which caused me to focus there. BUT I forgot that in a CoreXY machine, every transition affects every axis, X and Y are linked and must cooperate, so, even though the motion is in X, the hot end carriage is at the center of that motion.
When I dropped the hot end mount 10mm to give me space for a fourth slider, my binding issue got MUCH worse.I had the hot end mounted on a "side" slider instead of the "middle" one of three sliders, using three sliders that could not be helped since the belts attached to each side and other parts attached to the top "middle" one. This 10mm drop made the hot end carrier more unstable and I watched it flex, until it rolled over enough to bind.
Exoslide recommends that you mount the hot end on the middle of a three-sided slider or use four sliders.
I moved sliders around. Instead of three on each motion piece, I made one gantry slider a three, the other a two and made the hot end slider four. This required me to re-orient the "two slider" mount such that I could use the exact same mount as I use on the other end. This symmetry is good for a number of reasons and even though their orientation made the gantry mounts look asymmetric, it simplifies the design.
This three-two-four layout stopped binding (so far) and prints have even less ringing apparent. Everything seems to be much more stable. I will increase my speeds and see what happens, but it looks like this final change may have fixed my CoreXY problem.
More as I learn more,
DLC
Kits: Folgertech Kossel 2020 upgraded E3Dv6, Anet A8 upgraded E3Dv6, Tevo Tarantula enhanced parts and dual-head, TronXY X5SA Pro(E3DHemera).
Scratch: Large bed Cartesian, exchangeable heads, Linear slide Delta, Maker-Beam XL Micro Delta, 220x220CoreXY.